The present disclosure relates to computer software, and more specifically, to a framework for processing large datasets in partitioned distributed storage systems.
Distributed storage systems may enable large amounts of data to be stored in clusters of disparate compute nodes. Computer nodes in distributed storage system cluster may be arranged in a master-worker architecture with master nodes negotiating transactions with client applications and worker nodes executing these transactions. In some embodiments, data may be stored on compute nodes in partitions, or logical divisions of a storage space.